Question

Regression Analysis

Find a model from the 26 predictor variables using any model selection procedure you like to predict GIFTAMNT. GIFTAMNT is the amount of money donated by the donor in the last solicitation.

Also build a prediction model to predict NGIFTALL which helps us answer the question, that what leads people to gift more frequently? You must include NUMPRM12 and NUMPROM in this model to see on average how many times a donor needs to be contacted to generate a donation (i.e. a conversation rate of sorts). Conversion rate is of great interest to management as they want to identify donors which donate each time they are contacted. Getting small donations after contacting someone many of times may not be profitable as the donation might only cover the cost of postage of the numerous solicitation calls sent.

Please help me interpret the data.

SUMMARY OUTPUT Regression Statistics Multiple R 0.73730413 R Square Adjusted R Sq 0.54223669 Standard Errc 8.13870733 Observations 0.54361739 3648 ANOV df MS Significance F 0 Regression Residual Total 11 286879.148 26079.9226 393.72721 3636 240843.39366.238557 3647 527722.542 P-value Lower 95% Upper 95% Lower 99.0% Upper 99.0% 4.11220068 0.66003947 6.23023454 5.1891E-10 2.81811632 5.40628504 2.41115875 5.81324262 CARDPROM 0.03280286 0.07751233 0.42319544 0.6721777-0.1191691 0.184774830.16696053 0.23256626 NUMPROM 0.00556072 0.0311706 0.17839621 0.85842176-0.05555288 0.06667431 -0.07477159 0.08589302 0.1488036 0.17402378 0.85507628 0.39256526 -0.19239031 0.48999751-0.29968737 0.59729457 NUMPRM12 -0.0775417 0.0602889 -1.28616865 0.1984661 -0.19574513 0.04066173 -0.23291719 0.07783379 Coefficients Standard Error t Stat Intercept CARDPM12 0.06307885 0.04059849 -1.553723990.1203373 -0.14267692 0.01651923-0.16770855 0.04155086 CARDGIFT 0.12469982 0.07581755 -1.64473564 0.10011078 -0.27334897 0.02394932-0.32009546 0.07069581 MINRAMNT -0.18408597 0.03830115 -4.80627823 1.5999E-06 -0.25917984 -0.1089921 -0.282795-0.08537694 MAXRAMNT 0.00641124 0.00815369 0.78629935 0.43174337 -0.00957502 0.02239751-0.01460231 0.0274248 0.59890994 0.02076281 28.8453225 6.099E-165 0.55820203 0.63961785 0.5454004 0.65241948 0.37063018 0.04395482 8.43207126 4.8323E-17 0.28445162 0.456808730.2573506 0.48390975 CHILDREN -0.06154241 0.22992564 -0.26766222 0.78897452 -0.51233844 0.38925363-0.65410267 0.53101785 6 2 9 5 LASTGIFT AVGGIFT 2 7

Answer 1

Here we have given output of the regression model

From F statistics value is 393.727215 and its p value is zero

We know that if p value is less than 0 then model is significant Here p value less than alpha value hence regression model is significant..

Now for predictors here only MINRAMNT, LASTGIFT, AVGGIF

predictor have p value less than alpha value hence it predictor variable is significant and other are not significant